As health care delivery in the United States continues its transformation from an open, fee-for-service system into one dominated by capitation and the constraints of managed care, decisions about the availability of health services will be based increasingly on the results of medical outcome studies and cost-effectiveness analyses. There are few groups of patients for whom such decisions will have larger medical and economic impact then patients with osteoarthritis (OA). Moreover, given the tendency of medical outcomes studies to define health operationally as health-related functional status, few groups of patients will be more poorly served by such a bias than patients with OA--for whom relief of joint pain is paramount. The specific aims of this project will be (1) to establish a clinical routine for outpatient care of OA in which joint pain is monitored and charted as a vital sign in the medical record by an allied health professional and (2) to document the effects of longitudinal charting of such information on doctor-patient communication about pain, pain management practices of physicians, and patient outcomes (pain, satisfaction, and cost). We will utilize the practices of 9 rheumatologists (6 university- and 3 community-based) and a total of 90 patients with OA of the knee (10 patients/practice) in the study. For three consecutive office visits, each patient will undergo a pre- encounter assessment consisting of a validated measure of knee pain [the Western Ontario and McMaster Universities (WOMAC) OA Index], and ad hoc questionnaire measuring the patient's satisfaction with current control of joint pain, and a review of costs for arthritis drugs (NSAIDs, analgesics, anti-peptic-ulcer medications) since the last office visit. Following each office visit, a post-encounter interview will be held in which we will document any discussion of knee pain that took place between the patient and physician, any changes in prescription or over- the-counter (OTC) medications for knee OA, and the initiation of nonpharmacologic treatment (physical therapy, quadriceps-strengthening exercises, thermal modalities). Using a randomization scheme stratified for practice setting (university vs. community), the 9 rheumatologists (each with his or her cohort of 10 patients) will be assigned to one of three conditions: Pain-Charting (PC), cost-Monitoring (CM0, and pure control (CON). Under the PC condition, a graph will be appended to each patient's office record that will provide for the charting from visit to visit of WOMAC Pain scores [range: 5 (none) to 25 (extreme)] and the patient's concurrent satisfaction with control of knee pain. While it will have considerable intrinsic interest, the primary purpose of the CM group will be to serve as an attention-control condition. In the CM condition, a table containing a tally of the cumulative total and out-of- pocket costs to the patient for their OA drugs (NSAIDs, analgesics, anti- peptic ulcer drugs) will be appended to each patient's record. CON patients will receive the same pre- and post-encounter assessments as PC and CM patients, but no information will be brought to the attention of the physician for use during the visit. After the 10 patients in each practice have completed 3 office visits under their assigned condition PC, CM or CON), longitudinal and between-group comparisons will determine the effects of interventions on the frequency of doctor-patient communication about pain management, the prescribing practices of rheumatologists, and patient outcomes: joint pain, NSAID-induced side effects, satisfaction, and (controlling for medical insurance co-payment status) the cost of OA drugs.